Best student paper in EUSAR 2018

Our doctoral researcher Jian Kang won the 1st place of the student paper contest in EUSAR 2018, with the following contribution:

Kang J., Wang Y., Zhu X. (2018), Low Rank Modeling-based Multipass InSAR Technique. [paper]

A more detailed description of the award-winning algorithm can be found in the paper Object-Based Multipass InSAR via Robust Low-Rank Tensor Decomposition, IEEE TGRS, 56(6), 06.2018.

The following figure is a demonstration of the algorithm. It shows the linear deformation rates in mm/year (top) and elevation in meter (bottom) of Las Vegas Convention Center estimated by Persistent Scatterer Interferometry (PSI) and the proposed object-based multipass InSAR technique “RoMIO” using just nine TerraSAR-X high resolution spotlight images. As demonstrated in the results, large bias in the estimates is very likely to occur at low number of images or low SNR, due to the height ambiguity and the asymptotical optimality spectral estimator (i.e. periodogram) employed in PSI. In contrast, the proposed method can robustly recover the parameters with nearly no outliers using just nine interferograms.